Breakout Summary Report

Breakout Description

The DOE ARM user facility is establishing the third ARM Mobile Facility (AMF3) in the southeast United States for a five-year deployment starting in the fall of 2023, with siting focused on the Bankhead National Forest in northern Alabama. As site science team activities are beginning to shift from operations support to initial process studies, it is timely to begin more direct engagement with the Earth system model (ESM) community. The session focused on direct engagement with the weather, ESM, and LES communities -- strategizing ways to amplify the impact of the AMF3-BNF deployment as a next-generation model-observing system testbed to improve local to ESM model representations of coupled aerosol, cloud, and land-atmosphere interaction processes. Desired outcomes included: 1) discussing the community need for (and development of) flexible, computational frameworks for AMF3 model-observation integration (e.g., THREAD, LASSO); 2) defining observation and model needs and requirements by connecting modeling teams and experimentalists in order to develop data-model integration strategies for AMF3 and planned IOPs; 3) integrating observations across model systems that span broad spatio-temporal scales (e.g., WRF/LES, SCREAM-RRM, GCM); and 4) identifying new data tools, technologies, and infrastructure that are available (or needed) in order to support these integrative observation-modeling efforts.

Seven speakers were invited to present on modeling topics of their choice and how those topics could be pivoted towards the deployment (with an emphasis on approaches, challenges, and opportunities), across three themes: “Model-Observation Integration”, “Integrating Observations Across Scales”, and “Land-Atmosphere Interactions”.

• Girish Raghunathan, “Continuous Large-Eddy Simulations of the onset of convection over BNF and SGP”


• Manishkumar Shrivastava, ““SOA formation due to multiphase chemistry over the southeast USA: Implications from integrated model-measurements studies including WRF-Chem over the Amazon”


• Toshihisa Matsui and Marcus van Lier-Walqui, “Simulating isolated storms observed during TRACER using NU-WRF EPIC, a high-resolution weather model with polarimetric radar forward-simulation and prognostic electrification”


• Matt West, “Quantifying structural uncertainty in the aerosol modeling hierarchy: particle-resolved modeling on LES scales”


• Hsi-Yen Ma, “Regionally refined SCREAM for ARM sites”


• Marc Calaf, “A revised representation of atmospheric surface-layer processes over perturbed surfaces and the need for the next generation of field experiments”


• Thijs Heus, “Assessment of entrainment and advection in the diurnal cycle of the PBL over SGP and Bankhead National Forest”

Main Discussion

The main discussion centered primarily on needs, summarized here:

Measurement needs:

• Spatially resolved (e.g., AOS, aerosol node network, UAV, TBS), coordinated measurements of co-located oxidants, NOx, VOCs (seasonal -- cartridge-based; IOP - PTRMS), HOMs/ELVOCs/SVOCs/IVOCs, and high-resolution aerosol chemical composition to diagnose coupled biogenic emissions and SOA chemistry, and diurnal variability, as well as coupled controls of local sources/advection on aerosol spatial heterogeneity.


• Soil emissions (e.g., NOx, ammonia).


• Phenocams to capture ecosystem dynamics over time.


• High-spatial- (and volumetric), high-temporal-resolution radar observations to capture thermal evolution related to microphysics; agile and/or adaptive radar scanning (--> IOP).

Modeling needs:

• More detailed land-surface mapping (e.g., land cover type, vegetation) for model forcing.


• A consideration of running pre-deployment/ongoing LES/LASSO to inform siting/resiting of existing/additional observational resources.


• Critical setup specifics/considerations of radar simulators for measuring cloud fields.


• Model emulators to provide extended spatial data coverage, requiring training data and representative aerosol populations.


• Identification of BNF case studies for RRM-SCREAM to address convective cloud processes, land-atmosphere interactions (initially, testing with SGP/GoAmazon), and aerosol processes (one caveat: only prescribed aerosol in SCREAM currently)


•  High-resolution regional modeling constrained by the above measurements to understand the detailed processes governing new particle formation, SOA multiphase chemistry, cloud condensation nuclei and its connections with vegetation, land cover and clouds.

IOP needs:

• A thoughtful consideration of targeted time-scales (e.g., long-term, seasonal, sub-seasonal, diurnal). 


• A focus on more spatially extended flux observations to capture ecosystem scale fluxes, effects of canopy heterogeneity, and turbulence anisotropy.


• Advanced warning of IOP schedules (~ 1 year) to begin preliminary analysis and preparations.


• Readiness/availability of main site instrumentation (e.g., RWP, DL, MET) to capture impacts of solar eclipse (April 2024) on boundary-layer properties/processes.


• A campaign “dashboard” for campaign coordination, data visualization (e.g., quicklooks).


• Multi-agency/network/partner collaboration/coordination (e.g., mobile facilities/radars).


• A series of initial “shake-down” IOPs (across convective, aerosol, and land-atmosphere interactions science) to evaluate data quality and inform future, targeted measurements.


• An updated deployment website (e.g., planned instrument locations, community “Science Plan”).

Other Science Drivers or Considerations: 

• Impact of working forests on changing land use and local weather/climate, given the increasing demand for wood biomass/bioenergy in/from the southeast U.S.

Key Findings

Select key findings from the invited presentations:

• Automated LES-LSM model (e.g., MicroHH) compares well with surface SGP observations and ERA-5 reanalysis (over BNF), and will be installed on ARM Cumulus-II HPC for high-resolution simulations.


• In contrast to SGP, WRF-Chem simulations indicate IEPOX and terpene SOA dominate at the BNF, with potentially different roles of changing NOx, aqueous chemistry/cloud processing, and VOC sources.


• For convective cloud studies, radar simulators can play an important role in model-observing system integration when converting model-derived parameters for direct observational comparisons.


• Aerosol model structural uncertainty at the BNF can be driven by spatial heterogeneity (e.g., surface emission uncertainty/variability, 3D variability in aerosol size/composition) and aerosol composition heterogeneity (e.g., variability in internal/external particle population mixing state).


• Model-observing system integration of RRM-SCREAM for AMF3-BNF will initially require consideration of experimental domain design, land component configuration/resolution, and/or trade-offs of continuous runs with a smaller domain versus short runs (e.g., IOP mode) with a larger domain.


• Resolving the role of surface heterogeneity/non-local advective transport is critical to closing the surface energy balance, and can be realized by designing field experiments that “look” more like an LES control volume.


• Scalar (and possibly reactive species) profiles and fluxes can be reconstructed from LES output with reasonable accuracy, requiring only mean profiles, surface fluxes, and two sets of independent scalar observations.

Issues

None.

Needs

See main discussion above.

Decisions

None.

Future Plans

None.

Action Items

None.